A technique for training an equalizer of an ADSL Annex C transceiver is provided. The technique defines the points in the transmission sequence at which equalizer training should take place, and what symbols comprising the transmission should be used for training. The technique avoids the effect of strong TCM-ISDN cross-talk noise. The equalizer is trained so as to achieve optimal communication channel performance. During the TEQ training phase, the cyclic prefix is not introduced. The TEQ is trained only on FEXT symbols except the FEXT symbols adjacent to NEXT symbols. In the early phase of FEQ training, the cyclic prefix is not introduced. The FEQ is trained on FEXT symbols only except the FEXT symbols adjacent to NEXT symbols. During a late training phase after the cyclic prefix is introduced, the training of the FEQ is continued by using signals received on all FEXT symbols only. Due to the introduction of the cyclic prefix and the time-domain equalizer, the inter-symbol interference is basically eliminated. Therefore, the FEXT symbols adjacent to NEXT symbols are not affected by the TCM-ISDN noise that exists in NEXT symbols only. As a result, the received signal on those symbols can also be used to train FEQ. In a show-time phase, the FEQ is updated on FEXT and NEXT symbols for bins where the number of bits per symbol is non-zero.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A method for avoiding interference received by a Digital Subscriber Line (DSL) transceiver operating under a Time Compression Multiplexing-Integrated Services Digital Network (TCM-ISDN), wherein the transceiver includes a processor comprising a time domain equalizer and a number of frequency domain equalizers, the method comprising: selectively training the time-domain equalizer of the transceiver during a first training signal sequence where a first plurality of training symbols comprising FEXT and NEXT symbols are transmitted without a cyclic prefix to the transceiver; selectively training at least one of the frequency-domain equalizers of the transceiver during the first training signal sequence; and selectively training at least one of the frequency-domain equalizers of the transceiver during a second training signal sequence where a second plurality of training symbols comprising FEXT and NEXT symbols are transmitted with a cyclic prefix to the transceiver.
2. The method of claim 1 , wherein the first training signal sequence ranges from a transmission signal period C-PILOT 1 through a transmission signal period C-SEGUE 1 .
3. The method of claim 1 further comprising: before selectively training the time-domain equalizer, performing a TCM-ISDN timing reference alignment.
4. The method of claim 1 further comprising: adjusting an automatic gain control of the transceiver during the first training signal sequence.
5. The method of claim 4 , wherein the step of adjusting is performed during a transmission signal period C-REVERB 1 through a transmission signal period C-ECT.
6. The method of claim 1 , wherein the step of selectively training the time-domain equalizer is performed during a transmission signal period C-REVERB 2 .
7. The method of claim 1 , further comprising: aligning a symbol boundary of the first plurality of training symbols during a transmission signal period C-REVERB 2 .
8. The method of claim 1 , wherein the step of selectively training the time-domain equalizer is performed only under FEXT symbols that are not adjacent to NEXT symbols.
9. The method of claim 1 , wherein the step of selectively training at least one of the frequency-domain equalizers during the first training signal sequence is performed only under the FEXT symbols that are not adjacent to NEXT symbols.
10. The method of claim 1 , wherein the step of selectively training at least one of the frequency-domain equalizers during the second training signal sequence is performed under FEXT symbols, including a number of FEXT symbols adjacent to NEXT symbols.
11. The method of claim 1 , wherein the first plurality of training symbols further comprises a hyperframe, the hyperframe having 345 symbols sequentially numbered from left to right and top to bottom, the sequential numbering starting at symbol 0 in an upper left portion of the frame and ending at symbol 344 in a lower right portion of the frame.
12. The method of claim 11 , wherein the steps of selectively training the time-domain equalizer and selectively training at least one of the frequency-domain equalizers during the first training signal sequence are performed under any one of a subset of FEXT symbols included in the hyperframe, the subset of FEXT symbols including symbols: 1 , 2 , 12 , 13 , 23 , 24 , 34 , 35 , 44 , 45 , 55 , 56 , 66 , 67 , 77 , 78 , 88 , 89 , 98 , 99 , 109 , 110 , 120 , 121 , 131 , 132 , 141 , 142 , 143 , 152 , 153 , 163 , 164 , 174 , 175 , 185 , 186 , 195 , 196 , 206 , 207 , 217 , 218 , 228 , 229 , 238 , 239 , 240 , 249 , 250 , 260 , 261 , 271 , 272 , 282 , 283 , 292 , 293 , 303 , 304 , 314 , 315 , 325 , 326 , 336 , 337 .
13. The method of claim 11 , wherein the steps of selectively training the time-domain equalizer and selectively training at least one of the frequency-domain equalizers during the first training signal sequence are performed under any one of a subset of FEXT symbols included in the hyperframe, the subset of FEXT symbols including symbols: 7 , 8 , 17 , 18 , 28 , 29 , 39 , 40 , 50 , 51 , 61 , 62 , 71 , 72 , 82 , 83 , 93 , 94 , 104 , 105 , 114 , 115 , 116 , 125 , 126 , 136 , 137 , 147 , 148 , 158 , 159 , 168 , 169 , 179 , 180 , 190 , 191 , 201 , 202 , 211 , 212 , 213 , 222 , 223 , 233 , 234 , 244 , 245 , 255 , 256 , 265 , 266 , 276 , 277 , 287 , 288 , 298 , 299 , 309 , 310 , 319 , 320 , 330 , 331 , 341 , 342 .
14. The method of claim 1 , wherein the second training signal sequence ranges from a transmission signal period C-RATE 1 through a transmission signal period SHOWTIME.
15. The method of claim 1 , wherein the cyclic prefix is introduced during a transmission signal period C-RATE 1 .
16. The method of claim 1 further comprising: aligning a symbol boundary of the second plurality of training symbols during a transmission signal period C-RATE 1 .
17. The method of claim 1 , wherein the system operates in either a FEXT bitmap mode, or a dual bitmap mode.
18. The method of claim 1 , wherein the step of selectively training at least one of the frequency-domain equalizers during the second training signal sequence is performed under FEXT symbols, including a number of FEXT symbols adjacent to NEXT symbols, during a transmission signal period C-RATE 1 through a transmission signal period SHOWTIME.
19. The method of claim 1 , wherein the second plurality of training symbols further comprises a hyperframe, the hyperframe having 345 symbols sequentially numbered from left to right and top to bottom, the sequential numbering starting at symbol 0 in an upper left portion of the frame and ending at symbol 344 in a lower right portion of the frame.
20. The method of claim 19 , wherein the step of selectively training at least one of the frequency-domain equalizers during the second training signal sequence is performed under any one of a set of FEXT symbols included in the hyperframe, the set of FEXT symbols including symbols: 0 - 3 , 10 - 13 , 21 - 23 , 31 - 34 , 41 - 44 , 51 - 54 , 61 - 64 , 71 - 74 , 81 - 84 , 92 - 94 , 102 - 105 , 112 - 115 , 122 - 125 , 132 - 135 , 142 - 145 , 153 - 155 , 163 - 165 , 173 - 176 , 183 - 186 , 193 - 196 , 203 - 206 , 213 - 216 , 224 - 226 , 234 - 236 , 244 - 247 , 254 - 257 , 264 - 267 , 274 - 277 , 284 - 287 , 295 - 297 , 305 - 307 , 315 - 318 , 325 - 328 , 335 - 338 .
21. The method of claim 19 , wherein the step of selectively training at least one of the frequency-domain equalizers during the second training signal sequence is performed under any one of a set of FEXT symbols included in the hyperframe, the set of FEXT symbols including symbols: 5 - 8 , 15 - 18 , 26 - 28 , 36 - 39 , 46 - 49 , 56 - 59 , 66 - 69 76 - 79 , 87 - 89 , 97 - 99 , 107 - 110 , 117 - 120 , 127 - 130 , 137 - 140 , 147 - 150 , 158 - 160 , 168 - 170 , 178 - 181 , 188 - 191 , 198 - 201 , 208 - 211 , 218 - 221 , 229 - 231 , 239 - 241 , 249 - 252 , 259 - 262 , 269 - 272 , 279 - 282 , 289 - 292 , 300 - 302 , 310 - 313 , 320 - 323 , 330 - 333 , 340 - 343 .
22. The method of claim 1 , wherein the step of selectively training at least one of the frequency-domain equalizers during the first training signal sequence further comprises selectively updating a set of frequency-domain equalizer coefficients based on error signals generated by that frequency-domain equalizer and symbol decision.
23. The method of claim 1 , wherein the step of selectively training at least one of the frequency-domain equalizers during the second training signal sequence further comprises selectively updating a set of frequency-domain equalizer coefficients based on error signals generated by that frequency-domain equalizer and symbol decision.
24. The method of claim 1 further comprising: selectively updating at least one of the frequency-domain equalizers of the transceiver during the second training signal sequence.
25. The method of claim 24 wherein the step of selectively updating at least one of the frequency-domain equalizers further comprises selectively updating a set of frequency-domain equalizer coefficients based on error signals generated by that frequency-domain equalizer and symbol decision.
26. The method of claim 24 wherein the step of selectively updating at least one of the frequency-domain equalizers is performed under FEXT and NEXT symbols during a transmission signal period SHOWTIME.
27. The method of claim 24 , wherein the FEXT symbols are associated with a first bitmap and the NEXT symbols are associated with a second bitmap, the first and second bitmaps each being comprised of a number of sub-channels that are used for sending data, wherein the step of selectively updating at least one of the frequency-domain equalizers includes selectively updating a frequency-domain equalizer that is associated with a sub-channel that is used for sending data.
28. The method of claim 24 wherein the FEXT symbols are associated with a first bitmap and the NEXT symbols are associated with a second bitmap.
29. The method of claim 28 wherein each bitmap is comprised of a number of sub-channels that are used for sending data.
30. The method of claim 29 wherein each of the sub-channels that are used for sending data is associated with one of the frequency-domain equalizers.
31. The method of claim 30 wherein in response to one of the sub-channels being used to send data in NEXT time, but not in FEXT time, then the step of selectively updating at least one of the frequency-domain equalizers includes updating the frequency-domain equalizer associated with that sub-channel only in NEXT symbols.
32. The method of claim 30 wherein in response to one of the sub-channels being used to send data in FEXT time, but not in NEXT time, then the step of selectively updating at least one of the frequency-domain equalizers includes updating the frequency-domain equalizer associated with that sub-channel only in FEXT symbols.
33. The method of claim 30 wherein in response to one of the sub-channels being used to send data in NEXT time and FEXT time, then the step of selectively updating at least one of the frequency-domain equalizers includes updating the frequency-domain equalizer associated with that sub-channel in FEXT and NEXT symbols.
34. The method of claim 1 , wherein the first training signal sequence ranges from a transmission signal period R-REVERB 1 through a transmission signal period R-SEGUE 1 .
35. The method of claim 1 , wherein the step of selectively training the time-domain equalizer is performed during a transmission signal period R-REVERB 1 .
36. The method of claim 1 , further comprising: aligning a symbol boundary of the first plurality of training symbols during a transmission signal period R-REVERB 1 .
37. The method of claim 1 , wherein the second training signal sequence ranges from a transmission signal period R-REVERB 3 through a transmission signal period SHOWTIME.
38. The method of claim 1 , wherein the cyclic prefix is introduced during a transmission signal period R-REVERB 3 .
39. The method of claim 1 , further comprising: aligning a symbol boundary of the second plurality of training symbols during a transmission signal period R-REVERB 3 .
40. The method of claim 1 , wherein the step of selectively training at least one of the frequency-domain equalizers during the second training signal sequence is performed under all FEXT symbols during a transmission signal period R-REVERB 3 through a transmission signal period SHOWTIME.
41. A method for performing equalizer training for a Digital Subscriber Line (DSL) system operating under Time Compression Multiplexing-Integrated Services Digital Network (TCM-ISDN) interference, wherein the system includes a first transceiver and a second transceiver, each transceiver having a time-domain equalizer and a frequency-domain equalizer, the method comprising: transmitting a first symbol pattern comprising FEXT and NEXT symbols in a hyperframe without a cyclic prefix; training the time-domain equalizer of the second transceiver during the FEXT symbols of the first symbol pattern that are not adjacent to any of the NEXT symbols of the first symbol pattern; and training the frequency-domain equalizer of the second transceiver during the FEXT symbols of the first symbol pattern that are not adjacent to any of the NEXT symbols of the first symbol pattern.
42. The method of claim 41 , further comprising: transmitting a second symbol pattern comprising FEXT and NEXT symbols in a hyperframe with a cyclic prefix; and selectively training the frequency-domain equalizer of the second transceiver during the second symbol pattern.
43. A Digital Subscriber Line transceiver operating under TCM-ISDN interference, the transceiver comprising: a time-domain equalizer for compensating for channel distortion in time domain; a frequency-domain equalizer for compensating for channel distortion in frequency domain; and a digital signal processor for executing a process that comprises: training the time-domain equalizer during FEXT symbols of a first symbol pattern without a cyclic prefix, those FEXT symbols not being adjacent to any NEXT symbols of the first symbol pattern; and training the frequency-domain equalizer during the FEXT symbols of the first symbol pattern that are not adjacent to any of the NEXT symbols of the first symbol pattern.
44. The method of claim 43 , further comprising: selectively training the frequency-domain equalizer during a second symbol pattern with a cyclic prefix.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
February 29, 2000
September 30, 2003
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.